Comparative analysis for detecting objects under cast shadows in video images

Conference Article


International Conference on Pattern Recognition (ICPR)





Doc link


Download the digital copy of the doc pdf document


Cast shadows add additional difficulties on detecting objects because they locally modify image intensity and color. Shadows may appear or disappear in an image when the object, the camera, or both are free to move through a scene. In this work we use an object detection method based on boosted HOG paired with three different image representations, and we evaluate their relative performance. We follow and extend on the taxonomy from van de Sande [7] with considerations on the constraints assumed by each descriptor on the spatial variation of the illumination. We show that the intrinsic image representation consistently gives the best performance when tested on images from sequences acquired in an outdoor environment at different times of the day. This proves the usefulness of this representation for object detection in varying illumination conditions, and supports the idea that local assumptions in the descriptors can, in practice, be violated.


object detection.

Author keywords

image colour analysis , image representation , image sequences , lighting , object detection , video signal processing

Scientific reference

J. Scandaliaris, M. Villamizar and A. Sanfeliu. Comparative analysis for detecting objects under cast shadows in video images, 20th International Conference on Pattern Recognition, 2010, Istanbul, Turkey, pp. 4577-4580.